Matrix Vol. 12/1992, pp. 397 -403 © 1992 by Gustav Fischer Verlag, Stuttgart
An ELISA-Like Assay for Hyaluronidase and Hyaluronidase Inhibitors MICHAEL STERN 1 and ROBERT STERN 2 1
2
Departments of Stomatology and Oral and Maxillofacial Surgery, School of Dentistry and Department of Pathology, School of Medicine, University of California, San Francisco, CA 94143, USA.
Abstract
Hyaluronic acid (HA) is a prominent molecule in the extracellular matrix and is enriched whenever there is rapid tissue proliferation, regeneration and repair. HA is degraded in part by hyaluronidases (HA'ases) that are not well characterized. We have developed a novel ELISA-like rapid assay for HA'ases and their inhibitors. The assay is based on a high affinity biotinylated HA-binding peptide derived from tryptic digests of proteoglycan core protein of bovine nasal cartilage and the avidin-biotin reaction. HA-coated plates were incubated with serial dilutions of Streptomyces HA'ase, and the undegraded HA was measured. This established a standard curve for HA'ase activity against which all unknown enzyme samples were compared. The assay is easily modified to also serve a measure of HA'ase inhibitors. For detection of inhibitors, aliquots of sample were preincubated with a known activity of HA'ase and inhibition of HA degradation by the mixture was measured. We have used this assay to document the presence of potent HA'ase inhibitors in fetal calf sera. These techniques will aid in the purification and characterization ofHa'ases and their inhibitors. Key words: enzyme-linked immunosorbent assay, hyaluronic acid, hyaluronic acid-binding protein, hyaluronidase, hyaluronidase inhibitor.
Introduction
Hyaluronidases (HA'ases) are endoglycosidases that can hydrolyze the N-acetylglucosaminic bonds in hyaluronic acid (HA). Degradation of HA plays an important role in maintaining the integrity of the extracellular matrix. HA'ase and the degradation products of the reaction modulate such important biological processes as wound healing (Bertolami and Donoff, 1978; 1982; Thet et aI., 1983), angiogenesis (West et aI., 1985) and embryogenesis (Toole and Gross, 1971; Polansky et aI., 1973; Belsky and Toole, 1983; Kulyk and Kosher, 1987). Several assays have been established for the HA'ase group of enzymes. The original assays relied on the reduction of viscosity and turbidity of solutions containing HA (Dorfman, 1948; Dorfman and Ott, 1948) and activity was expressed as viscosity and turbidity reduction units. Currently, HA'ase activity is expressed in National Formulary
Units (NFU). The original techniques were useful in establishing sources rich in HA'ases but were relatively cumbersome and insensitive. The HA-impregnated agarose plate assay (Richman and Baer, 1980) is simple, requires only small sample volumes and is suitable for assaying multiple samples simultaneously. However, this technique is not as sensitive as other assays and has therefore not been widely used. More sensitive techniques include a dye-binding assay (Benchetrit et ai., 1977) and a recent assay that uses fluorogenic HA as substrate (Nakamura et aI., 1990). A zymogram electrophoretic technique using HA-impregnated polyacrylamide (Fiszer-Szafarz, 1984) has the advantage of separating HA'ase isoforms, and HA'ase from its inhibitors, but it is tedious and only semiquantitative. The Reissig assay (Reissig et aI., 1955) with a sensitivity of approximately 15 NFU (Linker, 1974), has been the most widely employed HA'ase assay. It is based upon generation of a new reducing N-acetylglucosamine terminus with each
398
M. Stern and R. Stern
cleavage reaction. Because this assay measures both terminal reducing N-acetylglucosamine and free N-acetylglucosamine, it is sensitive not only to the HA'ases, but also to the combined activities of ~-D-glucuronidase and Nacetyl-~-D-hexosaminidase. ~-D-glucuronidase attacks the nonreducing terminal glucuronic acid on the HA molecule. N-acetyl-~-D-hexosaminidase may then cleave off the nonreducing terminal N-acetylglucosamine resulting from the action of ~-D-glucuronidase. This reaction produces free N-acetylglucosamine that is detected by the Reissig assay. Both exoglycosidases are likely to be present in crude biological preparations, and the use of the Reissig assay may therefore give artifactually high levels ofHA'ase due to this activity. ~-glucuronidase can be specifically inhibited by including saccharolactone in the reaction mixture (Levvy and Marsh, 1959; Levvy and Conchie, 1966) and substitution of formate for acetate in the buffer results in its partial inhibition (Polansky et aI., 1973). A recent and sensitive HA'ase ELISA based on a brainderived HA-binding protein and antibodies against that molecule have been described (Delpech et aI., 1987). Despite its sensitivity, rapidity and convenience, this ELISA has also not been widely used, presumably due to the need for production and purification of hyaluronectin and antihyaluronectin antibodies. We report here a novel ELISAlike assay for HA'ase that is based on a cartilage-derived, biotinylated HA-binding protein and commercially available reagents. The assay can detect 1 x 10- 4 NFU of HA'ase and is rapid and simple. Multiple samples can be assayed simultaneously using small volumes of sample. The assay is 1,000 times more sensitive than the widely used Reissig assay. In addition, it can be modified easily to serve as an assay for HA'ase inhibitors. This latter group of substances, though of obvious importance in biological regulation, has not been well defined, presumably because of the lack of a sensitive, reproducible and rapid assay.
Materials and Methods Production ofHA-bindingprotein
The high affinity HA-binding protein was prepared according to Tengblad (1979), with some modifications, and then biotinylated (Bayer et aI., 1979). HA was coupled to AH-Sepharose (Pharmacia, Piscataway, NJ) as described (Tengblad, 1979). Approximately 150 g of bovine nasal cartilage (Pel Freeze, Rogers, AR) was stripped of perichondrium, diced, homogenized in 1 liter of ice cold 4 M guanidine hydrochloride, 0.5 M sodium acetate, pH 5.8, and then extracted for 24 h. This and all other steps in the HA-binding protein preparation were carried out at 4°C, except where indicated. The extract was passed over a Buchner funnel, without filter paper, to remove the largest cartilage fragments. The filtrate was centrifuged at 6000 g for 30 min and the supernatant collected and dialyzed
exhaustively against distilled water using Spectrapor-1 membranes (Spectrum, Los Angeles, CAl. A final dialysis was performed against 0.8 mM Tris HCl, 0.8 mM NaCI, pH 8.0. The dialyzed extract was lyophilized and stored at - 20°e. An aliquot (0.9 g) of lyophilized cartilage extract was rehydrated in 50 cc of buffer containing 0.1 M sodium acetate, 0.1 M Tris-HCI, pH 7.3. Tryptic peptides were generated by incubating with 2.0 mg of trypsin (type III; Sigma, StLouis, MO), at 37°e. After 2h, 1.2mg of soybean trypsin inhibitor (Worthington Chern. Co., Freehold, NJ) was added. This was then placed in dissociative conditions by bringing the extract to a concentration of 4 M guanidineHCl. This produces a disaggregation of the HA-binding peptide from HA. After 1 h, 70 ml of HA-Sepharose was added and the mixture dialyzed against water in order to "capture" the HA-binding region of the proteoglycan core protein with the HA-Sepharose. To facilitate optimal association of the proteoglycan core protein with the HASepharose, dialysis bags were inverted every 4-6 h to redistribute the settled gel. After exhaustive dialysis, a 70 ml bed volume column was prepared with the HA-Sepharose beads. To remove non-specifically adsorbed material, 300 ml each of 0.5 M sodium acetate with 1 M and 3 M NaCI, pH 5.8, were passed over the column at a flow rate of 26 mUh. The HA-binding peptides were then eluted from the column with 4 M guanidine-HCI, in 0.5 M sodium acetate, pH 5.8. Fractions of 5 ml were collected, and absorbance at 280 nm was recorded. Fractions containing the protein peak were pooled, concentrated in Centricon 10 tubes (Amicon, Beverly MA) to a final concentration of 1 mg/ml and dialyzed against 0.1 M sodium bicarbonate, pH 8.5. A total of 25 mg of protein was obtained. This HAbinding protein was then biotinylated with biotinyl Nhydroxysuccinimide ester according to the manufacturers instructions (Vector Laboratories, Burlingame, CAl, mixed with an equal volume of glycerol and stored at - 20 0e. Assay for hyaluronidase activity
The 96-well microtiter plates (Corning, Corning, NY) were coated with HA obtained from three commercial sources (Sigma, St. Louis, MO; ICN Biochemicals, Costa Mesa, CA; Pharmacia, Piscataway, NJ). This was performed to compare the ability of HA from three different sources to be used as substrates for this assay. The HA was dissolved in water at a concentration of 0.4 mg/ml. This was then diluted in an equal volume of 0.2 M carbonate buffer, pH 9.2. 100-I.ti aliquots of the diluted HA were applied to each well and the plate and incubated for 16 hat 4°e. To establish the ideal concentration of HA for adsorption to the microtiter plates, solutions ranging from 0.1-0.6 mg/ml HA in water were diluted in an equal volume of 0.2 M sodium carbonate buffer, pH 9.2. 100-I.ti
ELISA for Hyaluronidase and Inhibitors aliquots of the diluted HA were applied to each well. The plates were then incubated for 16 hours at 4°C and adsorbed HA detected as described below. All incubations were followed by three rinses in PBS containing 0.05% Tween 20 (Fisher Scientific, Fair Lawn, NJ). The wells were exposed to the wash buffer for approximately 15 s. All assays were performed in triplicate. Values represent the mean of triplicate wells; bars indicate the standard deviation. To establish concentration-dependence for the assay, the HA-coated wells were incubated with 100-111 aliquots for 5 h at 37°C with serial dilutions of Streptomyces HA'ase (Calbiochem, San Diego, CAl in 0.1 M sodium acetate, 0.15 M NaCl, 0.2 mg/ml BSA, pH 5.0. To establish timedependence, wells were incubated at 3l"C with 100-111 aliquots of 1 X 10- 3 NFU HA'ase for periods between 1 and 20h. Following the HA'ase incubation, non-specific binding by subsequent reagents was blocked by incubating with 300l1Vwell of ELISA blocking reagent (Boehringer Mannheim Biochemicals, Indianapolis, IN) for 30 min at 3 l"c. Alternatively, incubation with blocking reagent could be carried out prior to HA'ase digestion. The HA remaining after digestion was detected using the biotinylated HA-binding protein. To establish the ideal concentration of the HA-binding protein for detecting adsorbed HA, the protein was serially diluted in a buffer of 25 mM sodium phosphate, 0.15 M NaCl, 0.3 M guanidineHCl, 0.08% bovine serum albumin, and 0.02% sodium azide, pH 7.0. 100111 of the diluted HA-binding protein was then applied to each well and the plate was incubated for 1 h at 37°C. Next, to amplify the signal of biotinylation, wells were incubated with anti-keratan sulfate monoclonal antibody (ICN Biochemicals, Costa Mesa, CAl (1:1000 in PBS) for 30 min at room temperature. This was followed by incubating with biotinylated anti-mouse Ig (Vector, Burlingame, CAl (1:200 in PBS) for 30 min at room temperature. The biotinylated complex was detected with the avidinbiotin-peroxidase complex coupled to a reaction using 0phenylenediamine as a substrate as described by the manufacturer (Vector, Burlingame, CAl. Absorbance was read at 492nm. Streptomyces HA'ase was diluted in 0.1 M sodium acetate, 0.15 M NaCl, 0.2 mg/ml BSA. Liver HA'ase was prepared as described (Stern and Stern, 1990) and diluted in O.lM sodium formate, 0.15M NaCl, 0.2mg/ml BSA. To determine pH activity profiles, the pH of each enzyme was adjusted with acetic and formic acids, respectively. Protein concentrations were assayed using the BioRad protein dye kit with BSA used as a standard. To compare the sensitivity of the ELISA-like assay with an established method, we utilized the Reissig assay (1955). HA'ase activity was determined by measurement ofterminal N-acetylglucosamine released during incubation of HA with dilutions of Streptomyces HA'ase. 5-111 aliquots of
399
HA'ase were incubated at 3l"C for 16 h with 0.5 cc of 1 mg/ml HA and released N-acetylglucosamine measured colorimetrically as described.
Assay for HA'ase inhibitors in fetal calfserum The HA'ase assay was easily modified to serve as an assay for HA'ase inhibitors. Serial dilutions of fetal calf serum which is a potent source of inhibitor were made in PBS. 1 X 10- 3 U/ml of Streptomyces or partially purified liver HA'ase was then mixed with an equal volume of the serially diluted fetal calf serum or PBS. These mixtures of enzyme and inhibitor were incubated for 1 h at 37°C prior to application to the microtiter plate. During this incubation, inhibition of the enzyme by FCS occurred. 100-111 aliquots of these mixtures were then applied to HA coated wells and assayed for HA'ase activity as described above. The percent inhibition was calculated as %1 = 1 - [(Amax-Asample) I (A max - Amin )] where Amax is the absorbance of wells not exposed to HA'ase, Amin is the absorbance of wells exposed to HA'ase plus an equal volume of PBS (no inhibitor), and Asample is the absorbance of wells exposed to HA'ase plus an equal volume of sample containing inhibitor. Percent inhibition is thus determined from differences between the absence of HA degradation, HA degradation produced by a known enzymatic activity, and the degradation produced by a known enzyme activity exposed to enzyme inhibitors.
Results An idealized representation of the ELISA-like assay is presented in Figure 1. The proteoglycan core protein of cartilage from which the HA-binding protein is derived functions like an antibody to HA. The binding protein contains keratan sulfate glycosaminoglycan chains covalently attached to the core protein. We took advantage of this and amplified the signal of biotinylation on the HAbinding protein by exposing the HA-binding protein to an anti-keratan sulfate monoclonal antibody and a secondary biotinylated antibody (Fig. 1). Inclusion of the anti keratan sulfate antibody and biotinylated secondary antibody approximately doubled the sensitivity of the assay. The HA preparations from Sigma, ICN and Pharmacia were examined for use in this assay. Solutions of 0.4 mg/ml were diluted 1:2 in 0.2 M sodium carbonate buffer, pH 9.2 and incubated in the microtiter plate for 16 h at 4°C. The assay was run without exposure to HA'ase. HA from Sigma gave the highest level of responsiveness. This preparation of HA was therefore used as the substrate for all subsequent experiments. To determine the optimal concentration of HA needed to coat the microtiter plates, solutions of Sigma HA was prepared ranging from 100-60011g/ml in water and diluted 1:2 in sodium carbonate buffer to a final concentration
400
M. Stern and R. Stern Biotin
:::>-
:::>-Blotin
a KS
a mouse Ig
HASP
Fig. 1. Schematic diagram of hyaluronidase ELISA. Hyaluronic acid (HA) after HA'ase digestion was detected with the biotinylated HAbinding protein (HABP). The biotin signal was amplified by incubating with monoclonal antibody against keratan sulfate (aKS) and a secondary, biotinylated antibody. Bound complex was detected with the avidin-biotin-peroxidase technique.
0.75
0.8
0.7
E c
E c
-.
0.6
(II 0)
CIl
al
::!. «
0.65
0.5
c(
0.55
0.4
0.45 +--......--...,....--...--..,..--......--, 30 10 20 o
0.3
[HASP] [HAl J.lg/ml Fig. 2. Effect of HA concentration on responsiveness of the assay. Various concentrations of Sigma HA were diluted in water. These solutions were then diluted (1 :2) in carbonate buffer and adsorbed to wells as described in Material and Methods. [HA] values represent the HA concentration after dilution in carbonate buffer. In this and subsequent figures, values are means of triplicate wells; bars indicate standard deviation.
between 50-300 f..I.g/ml. The optimal concentration of HA for coating was 200 f..I.g/ml after dilution in carbonate buffer (Fig. 2). This concentration was therefore used to coat the wells in all subsequent experiments. The ideal concentration of the biotinylated HA-binding protein was determined by serial dilutions of the reagent in buffer. Again, the assay was performed without exposure to HA'ase to determine the maximal absorbance. Maximal absorbance was obtained when the HA-binding protein was diluted to a protein concentration of 5 f..I.g/ml (Fig. 3).
I!g/ml
Fig. 3. Effect of HA-binding protein concentration on the detection of adsorbed HA. To determine the minimal effective concentration of HA-binding protein needed for this ELISA, HA-binding protein was diluted in its buffer and used to detect adsorbed HA as described in Materials and Methods.
This concentration of the HA-binding protein was therefore used in all subsequent experiments. Enzyme assays must be both time- and dose-dependent to be valid. Such dependency is seen in Figures 4 and 5. HA degradation plateaus between 5 and 6 hours. The degradation is linear between 10- 3 and 10- 4 NFU/ml HA'ase. The degree of sensitivity of the ELISA-like assay is 1000 times greater than the Reissig assay. Figure 6 demonstrates a dose dependent decrease in N-acetylglucosamine released by dilutions of Streptomyces HA'ase. The limit of sensitivity of the Reissig assay in our hands is 0.1 NFU/ml HA'ase. A sensitivity of 15 NFU is reported by Linker (1974). Similar enzymatic activities from different sources can occasionally be distinguished by their pH profiles (Gold,
ELISA for Hyaluronidase and Inhibitors
401
12
>-
0.60
-
0.50
u
0.40
-
0.30
>
C
G»
~
u
.:
8
~
~
..
6
"t:l Cl
0.20
e Cl
Q)
a:
10
Ii
0.10
4
:>
au
0.00
oil:
0
5
15
10
Time
Z
20
til
:>
(hours)
0 .01
Fig. 4. Time-dependency of Streptomyces hyaluronidase enzyme activity. Hyaluronidase was added to HA-coated microtiter plates at a concentration of 1 x 10- 3 NFU and incubated for various times. Relative activity represents the difference in absorbance between wells not exposed to HA'ase and those exposed to this enzyme.
10
.1
100
1000
[HA'asel U/ml Fig. 6. Sensitivity of the Reissig assay. Dilutions of Streptomyces HA'ase were incubated with HA for 16 hours and released terminal N-acetylglucosamine measured as described. The assay can detect no less than 0.1 NFU/ml HA'ase. 0.5 0.4
E c
0.3
0.6
-
0.5
E
0.4
('II
0.3
c
m
2
4
6
8
[HA·...' U/ml x 10.
10
4
~
.c;
0.2 0.1 0.0
.01
.1
1
10
100 1000 [HA'ase] U/ml X 10· 4
10000
Fig. 5. Dose-dependency of Streptomyces hyaluronidase enzyme activity. Serial dilutions of hyaluronidase were incubated with HAcoated microtiter plates for 5 hours at 37°C as described in Materials and Methods. The reaction is linear between 10- 3 and 10- 4 U/ml (inset).
1982). We compared the activity at various pH's of Streptomyces HA'ase and a partially purified porcine liver HA'ase (Stern and Stern, 1990). Enzymes were diluted in buffer at the indicated pH's and assayed for activity (Fig. 7). Streptomyces HA'ase was most active at pH 5 and had a broad activity profile whereas the porcine liver HA'ase had
optimal activity at pH 4. The pH optima for the HA'ases found in porcine kidney extracts and the urine of children with Wilms' tumor are 3.5 (Stern et aI., 1991). By modifying the HA'ase assay to serve as a measure of HA'ase inhibitors, we were able to detect potent HA'ase inhibitors in fetal calf serum. Both Streptomyces and liver
402
M. Stern and R. Stern 0.5
The low sample volume and ability to assay multiple fractions make this assay particularly suitable for protein i.. 0.4 purification and for determination of activity in small ~ I ~ biological samples. The assay can detect as little as 1 x 10- 4 U 0.3 NFU of activity. This sensitivity is greater than that of any or: assay previously reported with the exception of the assay II Strep HA'••• 0.2 ~ described by Delpech (1987). The advantage of the assay ~ reported here is that all reagents are commercially available II 0.1 a:: or can be conveniently prepared. The HA-binding protein is LIver HA'aae easily purified and is a highly stable reagent. When stored at 0.0 - 20°C the binding protein retains its full activity for at 8 3 4 S 7 5 2 least 2.5 years. A second advantage of this assay is that it pH can be modified to serve as an assay for HA'ase inhibitors. Fig. 7. Activity of liver and Streptomyces HA'ases as a function of Serum contains HA'ase inhibitors and may playa role in pH. The Streptomyces enzyme was prepared in a buffer containing 0.1 M sodium acetate, 0.15 M NaCl 0.2 mg/ml BSA at the pHs tumor progression (Fiszer-Szafarz, 1968). However, these indicated. The liver HA'ase was prepared in a similar buffer con- inhibitors have not been isolated or characterized, primartaining sodium formate in place of sodium acetate. Relative activ- ily because a rapid and convenient assay has heretofore ity represents the difference in absorbance between wells not been lacking. exposed to HA'ase and those exposed to this enzyme. Since this assay uses substrate that is in solid phase, accurate determination of HA adsorption to the plate can not be made. This presents difficulties in calculations of 100 enzyme kinetics. In addition, cleavage of the HA polymer , by HA'ase produces "nicks" in the polymer that may not 80 necessarily result in displacement from the microtiter plate. We postulated that the negatively charged, hydrophilic HA so adsorbs to the negatively charged hydrophobic polystyrene plate primarily via HA-associated proteins. This suggests 40 that internal cleavage of the polymer may produce oligosac20 charide chains that are protein-free and therefore released from the plate. Regardless of the actual mechanism of HA displacement from the plate following exposure to HA'ase, ,01 ,1 the assay is time- and dose-dependent. FCS Dilution Regulation of HA metabolism in developing tissues, Fig. 8. FCS contains inhibitors of Streptomyces and liver HA'ases. healing wounds, and the stroma of malignant tumors plays FCS was serially diluted in PBS and incubated with an equal a critical role in these cellular processes. It is likely that volume of either Streptomyces or porcine liver HA'ase. Resulting regulation involves a balance between factors that stimulate enzymatic activity was then assayed as described in Materials and HA synthesis (Decker et aI., 1989) and factors that regulate Methods and activity compared to that in enzyme samples that its degradation. This degradation may in turn involve a had been incubated with PBS. balance between HA'ase and HA'ase inhibitors. Based on the information derived from the turnover of other HA'ases were inhibited by fetal calf serum in a dose depend- extracellular matrix molecules, we predict the existence of ent fashion (Fig. 8). Approximately 50% and 70% inhibi- several classes of physiologically relevant HA'ase tion was obtained by adding an equal volume of a 1:10 inhibitors. The ELISA-like assay described here will facilidilution of FCS to 1 X 10- 3 U/ml of Streptomyces and liver tate the purification and characterization of these imporHA'ase respectively. tant molecules.
,I
,,
,,
,i
Discussion We describe here a rapid and convenient ELISA-like assay for the detection of HA'ase and with minor modifications, an assay for HA'ase inhibitors. Multiple samples can be analyzed in less than 9 h. Since each well of the microtiter plate is incubated with 100 f..ll of sample, the total sample volume, tested in triplicate, requires no more than 300 f..lI.
Acknowledgements Supported by Public Health Service grant CA-44768 and HD 25505, National Cancer Institute, NIH, DHHS. M. Stern is supported by NIDR dentist-scientist award K15 DE00307-01. We thank Evangeline Leash for editorial assistance.
ELISA for Hyaluronidase and Inhibitors References Bayer, E.A., Skutelsky, E. and Wilchek, M.: The avidin-biotin complex in affinity cytochemistry. Methods Enzymol62: 308, 1979. Belsky, E. and Toole, B. P.: Hyaluronate and hyaluronidase in the developing chick embryo kidney. Cell Differentiation 12: 61-66, 1983. Benchetrit, L.c., Pahuja, S.L., Grey, E.D. and Edstrom, R.D.: A sensitive method for the assay of hyaluronidase activity. Anal. Biochem. 79: 431-437, 1977. Bertolami, C. N. and Donoff, R. B.: Hyaluronidase activity durin:g open wound healing in rabbits: A preliminary report. ]. Surg. Res. 25: 256-259, 1978. Bertolami, C. N. and Donoff, R. B.: Identification, characterization and partial purification of mammalian skin wound hyaluronidase.]. Invest. Dermatol. 79: 417 -421, 1982. Decker, M., Chiu, E.S., Dollbaum, C. Moiin, A., Hall, J., Spendlove, R., Longaker, M. and Stern, R.: Hyaluronic acid-stimulating activity in sera from the bovine fetus and from breast cancer patients. Cancer Res. 49: 3499-3505, 1989. Delpech, B., Bertrand, P. and Chauzy, c.: An indirect enzymoimmunological assay for hyaluronidase. ]. Immunol. Methods 104: 223-229, 1987. Dorfman, A.: The kinetics of the enzymatic hydrolysis of hyaluronic acid.]. Bioi. Chem. 172: 377, 1948. Dorfman, A. and Ott, M. L.: A turbidimetric method for the assay of hyaluronidase.]. BioI. Chem. 172: 367, 1948. Fiszer-Szafarz, B.: Demonstration of a new hyaluronidase inhibitor in serum of cancer patients. Proc. Soc. Exp. Bioi. Med. 12: 300-302, 1968. Fiszer-Szafarz, B.: Hyaluronidase polymorphism detected by polyacrylamide gel electrophoresis. Application to hyaluronidase from bacteria, slime molds, bee and snake venoms, bovine testes, rat liver lysosomes and human serum. Anal. Biochem. 143: 76-81, 1984. Gold, E. W.: Purification and properties of hyaluronidase from human liver. Differences from and similarities to the testicular enzyme. Biochem.]. 205: 69-74, 1982. Kulyk, W. M. and Kosher, R.A.: Temporal and spatial analysis of hyaluronidase activity during development of the embryonic chick limb bud. Dev. Bioi. 120: 535-541, 1987.
403
Levvy, G. A. and Marsh, A.: Preparation and properties of beta glucuronidase. Advan. Carbohyd. Chem. 14: 381-428, 1959. Levvy, G. A. and Conchie, J.: Mammalian glycosidases and their inhibition by aldonolactones. Method Enzymol. 8: 571-584, 1966. Linker, A.: Hyaluronidase. In: Methods of Enzymatic Analysis, Vol. 2, ed. by Bergmeyer, H. U., Academic Press, New York, 1974, pp. 944-948. Nakamura, T., Majima, M., Kubo, K., Takagahi, K., Tamura, S. and Endo, M.: Hyaluronidase assay using fluorogenic hyaluronate as substrate. Anal. Biochem. 191: 21-24, 1990. Polansky, J. R., Toole, B. P. and Gross, J.: Brain hyaluronidase: Changes in activity during chick development. Science 183: 862-863,1973. Reissig, J. L., Strominger, J. L. and Leloir, L. F.: A modified colorimetric method for the estimation ofN-acetylamino sugars.]. Bioi. Chem. 217: 959-966, 1955. Richman, P.G. and Baer, H.: A convenient plate assay for the quantitation of hyaluronidase in hymenoptera venoms. Anal. Biochem. 109: 376-381, 1980. Stern, M. and Stern, R.: Purification of hepatic hyaluronidase: Use of a novel ELISA-like assay.]. Cell Bioi. 111: 393 a, 1990. Stern, M., Longaker, M. T., Adzick, N. S., Harrison, M. and Stern, R.: Hyaluronidase levels in urine from Wilms' tumor patients.]. Natl. Cancer Inst. 83: 1569-1574,1991. Tengblad, A.: Affinity chromatography on immobilized hyaluronate and its application to the isolation of hyaluronate binding proteins from cartilage. Biochim. Biophys. Acta 578: 281-289,1979. Thet, L.A., Howell, A.C. and Han, G.: Changes in lung hyaluronidase activity associated with lung growth, injury and repair. Biochem. Biophys. Res. Commun. 117: 71-77, 1983. Toole, B. P. and Gross, J.: The extracellular matrix of the regenerating newt limb: Synthesis and removal of hyaluronate prior to differentiation. Dev. Bioi. 25: 57 -77, 1971. West, D.C., Hampson, LN., Arnold, F. and Kumar, S.: Angiogenesis induced by degradation products of hyaluronic acid. Science 228: 1324-1326,1985. Robert Stern, M.D., Department of Pathology, HSW-501, School of Medicine, University of California, San Francisco, CA 94143-0506, USA.
An ELISA-Like Assay for Hyaluronidase and Hyaluronidase Inhibitors MICHAEL STERN 1 and ROBERT STERN 2 1
2
Departments of Stomatology and Oral and Maxillofacial Surgery, School of Dentistry and Department of Pathology, School of Medicine, University of California, San Francisco, CA 94143, USA.
Abstract
Hyaluronic acid (HA) is a prominent molecule in the extracellular matrix and is enriched whenever there is rapid tissue proliferation, regeneration and repair. HA is degraded in part by hyaluronidases (HA'ases) that are not well characterized. We have developed a novel ELISA-like rapid assay for HA'ases and their inhibitors. The assay is based on a high affinity biotinylated HA-binding peptide derived from tryptic digests of proteoglycan core protein of bovine nasal cartilage and the avidin-biotin reaction. HA-coated plates were incubated with serial dilutions of Streptomyces HA'ase, and the undegraded HA was measured. This established a standard curve for HA'ase activity against which all unknown enzyme samples were compared. The assay is easily modified to also serve a measure of HA'ase inhibitors. For detection of inhibitors, aliquots of sample were preincubated with a known activity of HA'ase and inhibition of HA degradation by the mixture was measured. We have used this assay to document the presence of potent HA'ase inhibitors in fetal calf sera. These techniques will aid in the purification and characterization ofHa'ases and their inhibitors. Key words: enzyme-linked immunosorbent assay, hyaluronic acid, hyaluronic acid-binding protein, hyaluronidase, hyaluronidase inhibitor.
Introduction
Hyaluronidases (HA'ases) are endoglycosidases that can hydrolyze the N-acetylglucosaminic bonds in hyaluronic acid (HA). Degradation of HA plays an important role in maintaining the integrity of the extracellular matrix. HA'ase and the degradation products of the reaction modulate such important biological processes as wound healing (Bertolami and Donoff, 1978; 1982; Thet et aI., 1983), angiogenesis (West et aI., 1985) and embryogenesis (Toole and Gross, 1971; Polansky et aI., 1973; Belsky and Toole, 1983; Kulyk and Kosher, 1987). Several assays have been established for the HA'ase group of enzymes. The original assays relied on the reduction of viscosity and turbidity of solutions containing HA (Dorfman, 1948; Dorfman and Ott, 1948) and activity was expressed as viscosity and turbidity reduction units. Currently, HA'ase activity is expressed in National Formulary
Units (NFU). The original techniques were useful in establishing sources rich in HA'ases but were relatively cumbersome and insensitive. The HA-impregnated agarose plate assay (Richman and Baer, 1980) is simple, requires only small sample volumes and is suitable for assaying multiple samples simultaneously. However, this technique is not as sensitive as other assays and has therefore not been widely used. More sensitive techniques include a dye-binding assay (Benchetrit et ai., 1977) and a recent assay that uses fluorogenic HA as substrate (Nakamura et aI., 1990). A zymogram electrophoretic technique using HA-impregnated polyacrylamide (Fiszer-Szafarz, 1984) has the advantage of separating HA'ase isoforms, and HA'ase from its inhibitors, but it is tedious and only semiquantitative. The Reissig assay (Reissig et aI., 1955) with a sensitivity of approximately 15 NFU (Linker, 1974), has been the most widely employed HA'ase assay. It is based upon generation of a new reducing N-acetylglucosamine terminus with each
398
M. Stern and R. Stern
cleavage reaction. Because this assay measures both terminal reducing N-acetylglucosamine and free N-acetylglucosamine, it is sensitive not only to the HA'ases, but also to the combined activities of ~-D-glucuronidase and Nacetyl-~-D-hexosaminidase. ~-D-glucuronidase attacks the nonreducing terminal glucuronic acid on the HA molecule. N-acetyl-~-D-hexosaminidase may then cleave off the nonreducing terminal N-acetylglucosamine resulting from the action of ~-D-glucuronidase. This reaction produces free N-acetylglucosamine that is detected by the Reissig assay. Both exoglycosidases are likely to be present in crude biological preparations, and the use of the Reissig assay may therefore give artifactually high levels ofHA'ase due to this activity. ~-glucuronidase can be specifically inhibited by including saccharolactone in the reaction mixture (Levvy and Marsh, 1959; Levvy and Conchie, 1966) and substitution of formate for acetate in the buffer results in its partial inhibition (Polansky et aI., 1973). A recent and sensitive HA'ase ELISA based on a brainderived HA-binding protein and antibodies against that molecule have been described (Delpech et aI., 1987). Despite its sensitivity, rapidity and convenience, this ELISA has also not been widely used, presumably due to the need for production and purification of hyaluronectin and antihyaluronectin antibodies. We report here a novel ELISAlike assay for HA'ase that is based on a cartilage-derived, biotinylated HA-binding protein and commercially available reagents. The assay can detect 1 x 10- 4 NFU of HA'ase and is rapid and simple. Multiple samples can be assayed simultaneously using small volumes of sample. The assay is 1,000 times more sensitive than the widely used Reissig assay. In addition, it can be modified easily to serve as an assay for HA'ase inhibitors. This latter group of substances, though of obvious importance in biological regulation, has not been well defined, presumably because of the lack of a sensitive, reproducible and rapid assay.
Materials and Methods Production ofHA-bindingprotein
The high affinity HA-binding protein was prepared according to Tengblad (1979), with some modifications, and then biotinylated (Bayer et aI., 1979). HA was coupled to AH-Sepharose (Pharmacia, Piscataway, NJ) as described (Tengblad, 1979). Approximately 150 g of bovine nasal cartilage (Pel Freeze, Rogers, AR) was stripped of perichondrium, diced, homogenized in 1 liter of ice cold 4 M guanidine hydrochloride, 0.5 M sodium acetate, pH 5.8, and then extracted for 24 h. This and all other steps in the HA-binding protein preparation were carried out at 4°C, except where indicated. The extract was passed over a Buchner funnel, without filter paper, to remove the largest cartilage fragments. The filtrate was centrifuged at 6000 g for 30 min and the supernatant collected and dialyzed
exhaustively against distilled water using Spectrapor-1 membranes (Spectrum, Los Angeles, CAl. A final dialysis was performed against 0.8 mM Tris HCl, 0.8 mM NaCI, pH 8.0. The dialyzed extract was lyophilized and stored at - 20°e. An aliquot (0.9 g) of lyophilized cartilage extract was rehydrated in 50 cc of buffer containing 0.1 M sodium acetate, 0.1 M Tris-HCI, pH 7.3. Tryptic peptides were generated by incubating with 2.0 mg of trypsin (type III; Sigma, StLouis, MO), at 37°e. After 2h, 1.2mg of soybean trypsin inhibitor (Worthington Chern. Co., Freehold, NJ) was added. This was then placed in dissociative conditions by bringing the extract to a concentration of 4 M guanidineHCl. This produces a disaggregation of the HA-binding peptide from HA. After 1 h, 70 ml of HA-Sepharose was added and the mixture dialyzed against water in order to "capture" the HA-binding region of the proteoglycan core protein with the HA-Sepharose. To facilitate optimal association of the proteoglycan core protein with the HASepharose, dialysis bags were inverted every 4-6 h to redistribute the settled gel. After exhaustive dialysis, a 70 ml bed volume column was prepared with the HA-Sepharose beads. To remove non-specifically adsorbed material, 300 ml each of 0.5 M sodium acetate with 1 M and 3 M NaCI, pH 5.8, were passed over the column at a flow rate of 26 mUh. The HA-binding peptides were then eluted from the column with 4 M guanidine-HCI, in 0.5 M sodium acetate, pH 5.8. Fractions of 5 ml were collected, and absorbance at 280 nm was recorded. Fractions containing the protein peak were pooled, concentrated in Centricon 10 tubes (Amicon, Beverly MA) to a final concentration of 1 mg/ml and dialyzed against 0.1 M sodium bicarbonate, pH 8.5. A total of 25 mg of protein was obtained. This HAbinding protein was then biotinylated with biotinyl Nhydroxysuccinimide ester according to the manufacturers instructions (Vector Laboratories, Burlingame, CAl, mixed with an equal volume of glycerol and stored at - 20 0e. Assay for hyaluronidase activity
The 96-well microtiter plates (Corning, Corning, NY) were coated with HA obtained from three commercial sources (Sigma, St. Louis, MO; ICN Biochemicals, Costa Mesa, CA; Pharmacia, Piscataway, NJ). This was performed to compare the ability of HA from three different sources to be used as substrates for this assay. The HA was dissolved in water at a concentration of 0.4 mg/ml. This was then diluted in an equal volume of 0.2 M carbonate buffer, pH 9.2. 100-I.ti aliquots of the diluted HA were applied to each well and the plate and incubated for 16 hat 4°e. To establish the ideal concentration of HA for adsorption to the microtiter plates, solutions ranging from 0.1-0.6 mg/ml HA in water were diluted in an equal volume of 0.2 M sodium carbonate buffer, pH 9.2. 100-I.ti
ELISA for Hyaluronidase and Inhibitors aliquots of the diluted HA were applied to each well. The plates were then incubated for 16 hours at 4°C and adsorbed HA detected as described below. All incubations were followed by three rinses in PBS containing 0.05% Tween 20 (Fisher Scientific, Fair Lawn, NJ). The wells were exposed to the wash buffer for approximately 15 s. All assays were performed in triplicate. Values represent the mean of triplicate wells; bars indicate the standard deviation. To establish concentration-dependence for the assay, the HA-coated wells were incubated with 100-111 aliquots for 5 h at 37°C with serial dilutions of Streptomyces HA'ase (Calbiochem, San Diego, CAl in 0.1 M sodium acetate, 0.15 M NaCl, 0.2 mg/ml BSA, pH 5.0. To establish timedependence, wells were incubated at 3l"C with 100-111 aliquots of 1 X 10- 3 NFU HA'ase for periods between 1 and 20h. Following the HA'ase incubation, non-specific binding by subsequent reagents was blocked by incubating with 300l1Vwell of ELISA blocking reagent (Boehringer Mannheim Biochemicals, Indianapolis, IN) for 30 min at 3 l"c. Alternatively, incubation with blocking reagent could be carried out prior to HA'ase digestion. The HA remaining after digestion was detected using the biotinylated HA-binding protein. To establish the ideal concentration of the HA-binding protein for detecting adsorbed HA, the protein was serially diluted in a buffer of 25 mM sodium phosphate, 0.15 M NaCl, 0.3 M guanidineHCl, 0.08% bovine serum albumin, and 0.02% sodium azide, pH 7.0. 100111 of the diluted HA-binding protein was then applied to each well and the plate was incubated for 1 h at 37°C. Next, to amplify the signal of biotinylation, wells were incubated with anti-keratan sulfate monoclonal antibody (ICN Biochemicals, Costa Mesa, CAl (1:1000 in PBS) for 30 min at room temperature. This was followed by incubating with biotinylated anti-mouse Ig (Vector, Burlingame, CAl (1:200 in PBS) for 30 min at room temperature. The biotinylated complex was detected with the avidinbiotin-peroxidase complex coupled to a reaction using 0phenylenediamine as a substrate as described by the manufacturer (Vector, Burlingame, CAl. Absorbance was read at 492nm. Streptomyces HA'ase was diluted in 0.1 M sodium acetate, 0.15 M NaCl, 0.2 mg/ml BSA. Liver HA'ase was prepared as described (Stern and Stern, 1990) and diluted in O.lM sodium formate, 0.15M NaCl, 0.2mg/ml BSA. To determine pH activity profiles, the pH of each enzyme was adjusted with acetic and formic acids, respectively. Protein concentrations were assayed using the BioRad protein dye kit with BSA used as a standard. To compare the sensitivity of the ELISA-like assay with an established method, we utilized the Reissig assay (1955). HA'ase activity was determined by measurement ofterminal N-acetylglucosamine released during incubation of HA with dilutions of Streptomyces HA'ase. 5-111 aliquots of
399
HA'ase were incubated at 3l"C for 16 h with 0.5 cc of 1 mg/ml HA and released N-acetylglucosamine measured colorimetrically as described.
Assay for HA'ase inhibitors in fetal calfserum The HA'ase assay was easily modified to serve as an assay for HA'ase inhibitors. Serial dilutions of fetal calf serum which is a potent source of inhibitor were made in PBS. 1 X 10- 3 U/ml of Streptomyces or partially purified liver HA'ase was then mixed with an equal volume of the serially diluted fetal calf serum or PBS. These mixtures of enzyme and inhibitor were incubated for 1 h at 37°C prior to application to the microtiter plate. During this incubation, inhibition of the enzyme by FCS occurred. 100-111 aliquots of these mixtures were then applied to HA coated wells and assayed for HA'ase activity as described above. The percent inhibition was calculated as %1 = 1 - [(Amax-Asample) I (A max - Amin )] where Amax is the absorbance of wells not exposed to HA'ase, Amin is the absorbance of wells exposed to HA'ase plus an equal volume of PBS (no inhibitor), and Asample is the absorbance of wells exposed to HA'ase plus an equal volume of sample containing inhibitor. Percent inhibition is thus determined from differences between the absence of HA degradation, HA degradation produced by a known enzymatic activity, and the degradation produced by a known enzyme activity exposed to enzyme inhibitors.
Results An idealized representation of the ELISA-like assay is presented in Figure 1. The proteoglycan core protein of cartilage from which the HA-binding protein is derived functions like an antibody to HA. The binding protein contains keratan sulfate glycosaminoglycan chains covalently attached to the core protein. We took advantage of this and amplified the signal of biotinylation on the HAbinding protein by exposing the HA-binding protein to an anti-keratan sulfate monoclonal antibody and a secondary biotinylated antibody (Fig. 1). Inclusion of the anti keratan sulfate antibody and biotinylated secondary antibody approximately doubled the sensitivity of the assay. The HA preparations from Sigma, ICN and Pharmacia were examined for use in this assay. Solutions of 0.4 mg/ml were diluted 1:2 in 0.2 M sodium carbonate buffer, pH 9.2 and incubated in the microtiter plate for 16 h at 4°C. The assay was run without exposure to HA'ase. HA from Sigma gave the highest level of responsiveness. This preparation of HA was therefore used as the substrate for all subsequent experiments. To determine the optimal concentration of HA needed to coat the microtiter plates, solutions of Sigma HA was prepared ranging from 100-60011g/ml in water and diluted 1:2 in sodium carbonate buffer to a final concentration
400
M. Stern and R. Stern Biotin
:::>-
:::>-Blotin
a KS
a mouse Ig
HASP
Fig. 1. Schematic diagram of hyaluronidase ELISA. Hyaluronic acid (HA) after HA'ase digestion was detected with the biotinylated HAbinding protein (HABP). The biotin signal was amplified by incubating with monoclonal antibody against keratan sulfate (aKS) and a secondary, biotinylated antibody. Bound complex was detected with the avidin-biotin-peroxidase technique.
0.75
0.8
0.7
E c
E c
-.
0.6
(II 0)
CIl
al
::!. «
0.65
0.5
c(
0.55
0.4
0.45 +--......--...,....--...--..,..--......--, 30 10 20 o
0.3
[HASP] [HAl J.lg/ml Fig. 2. Effect of HA concentration on responsiveness of the assay. Various concentrations of Sigma HA were diluted in water. These solutions were then diluted (1 :2) in carbonate buffer and adsorbed to wells as described in Material and Methods. [HA] values represent the HA concentration after dilution in carbonate buffer. In this and subsequent figures, values are means of triplicate wells; bars indicate standard deviation.
between 50-300 f..I.g/ml. The optimal concentration of HA for coating was 200 f..I.g/ml after dilution in carbonate buffer (Fig. 2). This concentration was therefore used to coat the wells in all subsequent experiments. The ideal concentration of the biotinylated HA-binding protein was determined by serial dilutions of the reagent in buffer. Again, the assay was performed without exposure to HA'ase to determine the maximal absorbance. Maximal absorbance was obtained when the HA-binding protein was diluted to a protein concentration of 5 f..I.g/ml (Fig. 3).
I!g/ml
Fig. 3. Effect of HA-binding protein concentration on the detection of adsorbed HA. To determine the minimal effective concentration of HA-binding protein needed for this ELISA, HA-binding protein was diluted in its buffer and used to detect adsorbed HA as described in Materials and Methods.
This concentration of the HA-binding protein was therefore used in all subsequent experiments. Enzyme assays must be both time- and dose-dependent to be valid. Such dependency is seen in Figures 4 and 5. HA degradation plateaus between 5 and 6 hours. The degradation is linear between 10- 3 and 10- 4 NFU/ml HA'ase. The degree of sensitivity of the ELISA-like assay is 1000 times greater than the Reissig assay. Figure 6 demonstrates a dose dependent decrease in N-acetylglucosamine released by dilutions of Streptomyces HA'ase. The limit of sensitivity of the Reissig assay in our hands is 0.1 NFU/ml HA'ase. A sensitivity of 15 NFU is reported by Linker (1974). Similar enzymatic activities from different sources can occasionally be distinguished by their pH profiles (Gold,
ELISA for Hyaluronidase and Inhibitors
401
12
>-
0.60
-
0.50
u
0.40
-
0.30
>
C
G»
~
u
.:
8
~
~
..
6
"t:l Cl
0.20
e Cl
Q)
a:
10
Ii
0.10
4
:>
au
0.00
oil:
0
5
15
10
Time
Z
20
til
:>
(hours)
0 .01
Fig. 4. Time-dependency of Streptomyces hyaluronidase enzyme activity. Hyaluronidase was added to HA-coated microtiter plates at a concentration of 1 x 10- 3 NFU and incubated for various times. Relative activity represents the difference in absorbance between wells not exposed to HA'ase and those exposed to this enzyme.
10
.1
100
1000
[HA'asel U/ml Fig. 6. Sensitivity of the Reissig assay. Dilutions of Streptomyces HA'ase were incubated with HA for 16 hours and released terminal N-acetylglucosamine measured as described. The assay can detect no less than 0.1 NFU/ml HA'ase. 0.5 0.4
E c
0.3
0.6
-
0.5
E
0.4
('II
0.3
c
m
2
4
6
8
[HA·...' U/ml x 10.
10
4
~
.c;
0.2 0.1 0.0
.01
.1
1
10
100 1000 [HA'ase] U/ml X 10· 4
10000
Fig. 5. Dose-dependency of Streptomyces hyaluronidase enzyme activity. Serial dilutions of hyaluronidase were incubated with HAcoated microtiter plates for 5 hours at 37°C as described in Materials and Methods. The reaction is linear between 10- 3 and 10- 4 U/ml (inset).
1982). We compared the activity at various pH's of Streptomyces HA'ase and a partially purified porcine liver HA'ase (Stern and Stern, 1990). Enzymes were diluted in buffer at the indicated pH's and assayed for activity (Fig. 7). Streptomyces HA'ase was most active at pH 5 and had a broad activity profile whereas the porcine liver HA'ase had
optimal activity at pH 4. The pH optima for the HA'ases found in porcine kidney extracts and the urine of children with Wilms' tumor are 3.5 (Stern et aI., 1991). By modifying the HA'ase assay to serve as a measure of HA'ase inhibitors, we were able to detect potent HA'ase inhibitors in fetal calf serum. Both Streptomyces and liver
402
M. Stern and R. Stern 0.5
The low sample volume and ability to assay multiple fractions make this assay particularly suitable for protein i.. 0.4 purification and for determination of activity in small ~ I ~ biological samples. The assay can detect as little as 1 x 10- 4 U 0.3 NFU of activity. This sensitivity is greater than that of any or: assay previously reported with the exception of the assay II Strep HA'••• 0.2 ~ described by Delpech (1987). The advantage of the assay ~ reported here is that all reagents are commercially available II 0.1 a:: or can be conveniently prepared. The HA-binding protein is LIver HA'aae easily purified and is a highly stable reagent. When stored at 0.0 - 20°C the binding protein retains its full activity for at 8 3 4 S 7 5 2 least 2.5 years. A second advantage of this assay is that it pH can be modified to serve as an assay for HA'ase inhibitors. Fig. 7. Activity of liver and Streptomyces HA'ases as a function of Serum contains HA'ase inhibitors and may playa role in pH. The Streptomyces enzyme was prepared in a buffer containing 0.1 M sodium acetate, 0.15 M NaCl 0.2 mg/ml BSA at the pHs tumor progression (Fiszer-Szafarz, 1968). However, these indicated. The liver HA'ase was prepared in a similar buffer con- inhibitors have not been isolated or characterized, primartaining sodium formate in place of sodium acetate. Relative activ- ily because a rapid and convenient assay has heretofore ity represents the difference in absorbance between wells not been lacking. exposed to HA'ase and those exposed to this enzyme. Since this assay uses substrate that is in solid phase, accurate determination of HA adsorption to the plate can not be made. This presents difficulties in calculations of 100 enzyme kinetics. In addition, cleavage of the HA polymer , by HA'ase produces "nicks" in the polymer that may not 80 necessarily result in displacement from the microtiter plate. We postulated that the negatively charged, hydrophilic HA so adsorbs to the negatively charged hydrophobic polystyrene plate primarily via HA-associated proteins. This suggests 40 that internal cleavage of the polymer may produce oligosac20 charide chains that are protein-free and therefore released from the plate. Regardless of the actual mechanism of HA displacement from the plate following exposure to HA'ase, ,01 ,1 the assay is time- and dose-dependent. FCS Dilution Regulation of HA metabolism in developing tissues, Fig. 8. FCS contains inhibitors of Streptomyces and liver HA'ases. healing wounds, and the stroma of malignant tumors plays FCS was serially diluted in PBS and incubated with an equal a critical role in these cellular processes. It is likely that volume of either Streptomyces or porcine liver HA'ase. Resulting regulation involves a balance between factors that stimulate enzymatic activity was then assayed as described in Materials and HA synthesis (Decker et aI., 1989) and factors that regulate Methods and activity compared to that in enzyme samples that its degradation. This degradation may in turn involve a had been incubated with PBS. balance between HA'ase and HA'ase inhibitors. Based on the information derived from the turnover of other HA'ases were inhibited by fetal calf serum in a dose depend- extracellular matrix molecules, we predict the existence of ent fashion (Fig. 8). Approximately 50% and 70% inhibi- several classes of physiologically relevant HA'ase tion was obtained by adding an equal volume of a 1:10 inhibitors. The ELISA-like assay described here will facilidilution of FCS to 1 X 10- 3 U/ml of Streptomyces and liver tate the purification and characterization of these imporHA'ase respectively. tant molecules.
,I
,,
,,
,i
Discussion We describe here a rapid and convenient ELISA-like assay for the detection of HA'ase and with minor modifications, an assay for HA'ase inhibitors. Multiple samples can be analyzed in less than 9 h. Since each well of the microtiter plate is incubated with 100 f..ll of sample, the total sample volume, tested in triplicate, requires no more than 300 f..lI.
Acknowledgements Supported by Public Health Service grant CA-44768 and HD 25505, National Cancer Institute, NIH, DHHS. M. Stern is supported by NIDR dentist-scientist award K15 DE00307-01. We thank Evangeline Leash for editorial assistance.
ELISA for Hyaluronidase and Inhibitors References Bayer, E.A., Skutelsky, E. and Wilchek, M.: The avidin-biotin complex in affinity cytochemistry. Methods Enzymol62: 308, 1979. Belsky, E. and Toole, B. P.: Hyaluronate and hyaluronidase in the developing chick embryo kidney. Cell Differentiation 12: 61-66, 1983. Benchetrit, L.c., Pahuja, S.L., Grey, E.D. and Edstrom, R.D.: A sensitive method for the assay of hyaluronidase activity. Anal. Biochem. 79: 431-437, 1977. Bertolami, C. N. and Donoff, R. B.: Hyaluronidase activity durin:g open wound healing in rabbits: A preliminary report. ]. Surg. Res. 25: 256-259, 1978. Bertolami, C. N. and Donoff, R. B.: Identification, characterization and partial purification of mammalian skin wound hyaluronidase.]. Invest. Dermatol. 79: 417 -421, 1982. Decker, M., Chiu, E.S., Dollbaum, C. Moiin, A., Hall, J., Spendlove, R., Longaker, M. and Stern, R.: Hyaluronic acid-stimulating activity in sera from the bovine fetus and from breast cancer patients. Cancer Res. 49: 3499-3505, 1989. Delpech, B., Bertrand, P. and Chauzy, c.: An indirect enzymoimmunological assay for hyaluronidase. ]. Immunol. Methods 104: 223-229, 1987. Dorfman, A.: The kinetics of the enzymatic hydrolysis of hyaluronic acid.]. Bioi. Chem. 172: 377, 1948. Dorfman, A. and Ott, M. L.: A turbidimetric method for the assay of hyaluronidase.]. BioI. Chem. 172: 367, 1948. Fiszer-Szafarz, B.: Demonstration of a new hyaluronidase inhibitor in serum of cancer patients. Proc. Soc. Exp. Bioi. Med. 12: 300-302, 1968. Fiszer-Szafarz, B.: Hyaluronidase polymorphism detected by polyacrylamide gel electrophoresis. Application to hyaluronidase from bacteria, slime molds, bee and snake venoms, bovine testes, rat liver lysosomes and human serum. Anal. Biochem. 143: 76-81, 1984. Gold, E. W.: Purification and properties of hyaluronidase from human liver. Differences from and similarities to the testicular enzyme. Biochem.]. 205: 69-74, 1982. Kulyk, W. M. and Kosher, R.A.: Temporal and spatial analysis of hyaluronidase activity during development of the embryonic chick limb bud. Dev. Bioi. 120: 535-541, 1987.
403
Levvy, G. A. and Marsh, A.: Preparation and properties of beta glucuronidase. Advan. Carbohyd. Chem. 14: 381-428, 1959. Levvy, G. A. and Conchie, J.: Mammalian glycosidases and their inhibition by aldonolactones. Method Enzymol. 8: 571-584, 1966. Linker, A.: Hyaluronidase. In: Methods of Enzymatic Analysis, Vol. 2, ed. by Bergmeyer, H. U., Academic Press, New York, 1974, pp. 944-948. Nakamura, T., Majima, M., Kubo, K., Takagahi, K., Tamura, S. and Endo, M.: Hyaluronidase assay using fluorogenic hyaluronate as substrate. Anal. Biochem. 191: 21-24, 1990. Polansky, J. R., Toole, B. P. and Gross, J.: Brain hyaluronidase: Changes in activity during chick development. Science 183: 862-863,1973. Reissig, J. L., Strominger, J. L. and Leloir, L. F.: A modified colorimetric method for the estimation ofN-acetylamino sugars.]. Bioi. Chem. 217: 959-966, 1955. Richman, P.G. and Baer, H.: A convenient plate assay for the quantitation of hyaluronidase in hymenoptera venoms. Anal. Biochem. 109: 376-381, 1980. Stern, M. and Stern, R.: Purification of hepatic hyaluronidase: Use of a novel ELISA-like assay.]. Cell Bioi. 111: 393 a, 1990. Stern, M., Longaker, M. T., Adzick, N. S., Harrison, M. and Stern, R.: Hyaluronidase levels in urine from Wilms' tumor patients.]. Natl. Cancer Inst. 83: 1569-1574,1991. Tengblad, A.: Affinity chromatography on immobilized hyaluronate and its application to the isolation of hyaluronate binding proteins from cartilage. Biochim. Biophys. Acta 578: 281-289,1979. Thet, L.A., Howell, A.C. and Han, G.: Changes in lung hyaluronidase activity associated with lung growth, injury and repair. Biochem. Biophys. Res. Commun. 117: 71-77, 1983. Toole, B. P. and Gross, J.: The extracellular matrix of the regenerating newt limb: Synthesis and removal of hyaluronate prior to differentiation. Dev. Bioi. 25: 57 -77, 1971. West, D.C., Hampson, LN., Arnold, F. and Kumar, S.: Angiogenesis induced by degradation products of hyaluronic acid. Science 228: 1324-1326,1985. Robert Stern, M.D., Department of Pathology, HSW-501, School of Medicine, University of California, San Francisco, CA 94143-0506, USA.
